Hi,
This is the position (Black to move):
I use it regularly to test new engines, see how fast they can solve it. It's still a mystery to me that even today's most advanced engines can't solve it right away (except Dragon).
The ONLY engine that can solve it immediately (on my machine) is Dragon (formerly Komodo). It solves it in half a second (or less), using the AVX2 version obviously.
Even the mighty Stockfish (version 13 tested today) takes a few seconds. (Sf 13 is a noticeable improvement over version 12 which took almost 20 seconds!)
My pc specs:
CPU: Intel i9 10900K (10 cores @4.9ghz)
RAM: gSkill 32gb (2x16gb) 3600mhz CL16
SSD: 1TB Samsung 970 evo plus (nvme)
Mainboard: Gigabyte Z490 Aorus Master
I use Fritz17 GUI with 8gb of hash (without syzygy TB, but when I tested with Syzygy on it didn't make a difference [strangely])
How fast can your engine see the solution to this position?
Moderators: hgm, Rebel, chrisw
-
- Posts: 20
- Joined: Fri Sep 04, 2020 9:20 pm
- Full name: Assaf Patishi
-
- Posts: 1142
- Joined: Thu Dec 28, 2017 4:06 pm
- Location: Argentina
Re: How fast can your engine see the solution to this position?
And what is the solution? Can you provide FEN?
Follow my tournament and some Leela gauntlets live at http://twitch.tv/ccls
-
- Posts: 20
- Joined: Fri Sep 04, 2020 9:20 pm
- Full name: Assaf Patishi
Re: How fast can your engine see the solution to this position?
Well, this is a known position from Topalov vs. Shirov (Linares 1998). I thought it would be interesting if you solve it with your engines (or by yourself if you can) But anyway.. Black only move to win is 47..Bh3.
I don't know how to provide FEN
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: How fast can your engine see the solution to this position?
It's the position from Shirov's Bh3! : http://www.talkchess.com/forum3/viewtop ... =2&t=44282
-
- Posts: 16
- Joined: Thu Feb 18, 2021 10:06 pm
- Full name: Lev Chen
Re: How fast can your engine see the solution to this position?
[d]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 24
This is a better position to test Dragon Komodo, how long does it take to solve it?
Best move is 1.Qxe5!.
Crystal 3.1 finds the move right away!!!
This is a better position to test Dragon Komodo, how long does it take to solve it?
Best move is 1.Qxe5!.
Crystal 3.1 finds the move right away!!!
-
- Posts: 20
- Joined: Fri Sep 04, 2020 9:20 pm
- Full name: Assaf Patishi
Re: How fast can your engine see the solution to this position?
Very nice position! Thx. It is interesting that the best playing engines like Stockfish is not necessarily better at solving certain positions.NoChessNoLife wrote: ↑Sat Feb 20, 2021 3:45 am [d]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 24
This is a better position to test Dragon Komodo, how long does it take to solve it?
Best move is 1.Qxe5!.
Crystal 3.1 finds the move right away!!!
Dragon sees the solution faster than Stockfish 13 but still it takes time, not right away..not even close.
-
- Posts: 20
- Joined: Fri Sep 04, 2020 9:20 pm
- Full name: Assaf Patishi
Re: How fast can your engine see the solution to this position?
For some reason, I can't edit my posts.. strange.
But I want to update about something weird I found, when in multi vaiation mode (I usually use 2 when analyzing) Dragon find Bh3 faster than stockfish but when in single variation, it takes pretty long time. While stockfish sees the solution pretty quickly.
I think at the moment I will use Stockfish 13 for analyzing cause it seems to be the strongest and most stable.
But I want to update about something weird I found, when in multi vaiation mode (I usually use 2 when analyzing) Dragon find Bh3 faster than stockfish but when in single variation, it takes pretty long time. While stockfish sees the solution pretty quickly.
I think at the moment I will use Stockfish 13 for analyzing cause it seems to be the strongest and most stable.
-
- Posts: 584
- Joined: Fri Mar 30, 2018 7:20 am
- Full name: Andreas Matthies
Re: How fast can your engine see the solution to this position?
My engine finds Bh3 at depth 61 after 16:31min. Ryzen 9 4900H, one thread, 2GB Hash.
In comparison SF12 on the same machine and conditions finds it at depth 57 after 1:39min.
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: How fast can your engine see the solution to this position?
With partial 7-piece WDL on M.2 SSD. It takes SF 13 just 3-4 seconds to spot Bh3 on my aging 5960X at 4.5GHz.
-
- Posts: 1968
- Joined: Wed Jul 13, 2011 9:04 pm
- Location: Madrid, Spain.
Re: How fast can your engine see the solution to this position?
Hello:
My post is not about how my HW performs in this gem but about bringing back a piece of history: the then (and currently) best-selling newspaper in Spain, El País, has a chess column. This famous game was played on March 4th, 1998 in Linares (Jaén, Spain) and the next issue of El País dedicated an article in its sports section to what happened the day before at Linares chess tournament and specially in this game. The article can be read in Spanish in the following link:
Shírov produce arte y vuelve a dominar
Online translators can help you. It was the fifth hour of the match and all the GMs out there were analyzing the position. 30 minutes later they were agreeing about the game outcome, which should be a draw. Fritz 5 confirmed GMs thoughts, when Shirov played 47.- ..., Bh3!! and first impressions were 'Blunder!', until someone spotted: 'If Shirov has played that, there must me a plan'.
Interesting story of the first opinions on the move and the use of Fritz 5 no less. Computer chess has improved a lot since then and has to improve Elo-wise much more in the following decades.
Regards from Spain.
Ajedrecista.
My post is not about how my HW performs in this gem but about bringing back a piece of history: the then (and currently) best-selling newspaper in Spain, El País, has a chess column. This famous game was played on March 4th, 1998 in Linares (Jaén, Spain) and the next issue of El País dedicated an article in its sports section to what happened the day before at Linares chess tournament and specially in this game. The article can be read in Spanish in the following link:
Shírov produce arte y vuelve a dominar
Online translators can help you. It was the fifth hour of the match and all the GMs out there were analyzing the position. 30 minutes later they were agreeing about the game outcome, which should be a draw. Fritz 5 confirmed GMs thoughts, when Shirov played 47.- ..., Bh3!! and first impressions were 'Blunder!', until someone spotted: 'If Shirov has played that, there must me a plan'.
Interesting story of the first opinions on the move and the use of Fritz 5 no less. Computer chess has improved a lot since then and has to improve Elo-wise much more in the following decades.
Regards from Spain.
Ajedrecista.